INDEX

Explanations

references to image credits.

oai_token-act-pair · gpt-3.5-turbo

references to image credits or attributions in media content

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

etime

-0.80

stood

-0.77

 disarm

-0.72

ced

-0.71

gone

-0.70

cible

-0.70

 congress

-0.70

ggle

-0.68

hillary

-0.66

turn

-0.65

POSITIVE LOGITS

 Images

1.25

 Thumbnails

1.02

Images

0.94

 Photography

0.92

 Photo

0.86

Photos

0.83

 Gallery

0.83

 caption

0.82

 Photographer

0.82

 IMAGES

0.82

Activations Density 0.015%

references to image credits.

references to image credits or attributions in media content

No Comments

No Known Activations